Did it work, and was the engine right?
Every decision is measured against stores where it didn’t run, so you see what was really added. Then you see which levers pay best for your business, and where the engine’s forecasts can be trusted most.
Answers, not report builders.
Promotion results
Daily sales where a promotion ran against the holdout group, forecast against actual, and the full cost breakdown.
ExampleDid it add sales, or would they have happened anyway?
Levers and decisions
Return for every ₹1 by lever and offer type, and forecast accuracy by lever.
ExampleWhich levers work best for this business?
Engine vs your own picks
Recommended, second and client’s own options side by side, plus reasons recommendations were turned down.
ExampleDo the engine’s picks beat ours?
Partners
Paid to partners, extra units, cost per extra unit against the markdown alternative, money waiting on return windows.
ExampleWhich partners to book again?
SKU health over time
Dead stock, ageing and stock-outs by category, week by week.
ExampleAre we carrying less trouble than last season?
Data feeds and runs
Last run, data health over 30 days, every feed’s status and the rows set aside, by reason. Automatic decisions pause on their own when health drops below your threshold.
ExampleDid the engine run properly today?
09Walkthrough on your data
Bring one category. See tomorrow’s worklist.
We load a sample of your catalogue, sales and stock, run the engine on it, and walk you through every recommendation, the reason behind it and what it would have earned.
Go live with as few as 100 SKUs and any length of sales history.
- Your catalogue and stock, loaded
- SKU health and risk scores
- Buy, move and markdown calls
- Festival effects for your regions
- Your brand and store rules
- Live cart or recommendations only
- What it would have saved
- A rollout plan by category